Accidents between right-turning vehicles and straight-ahead non-motor vehicles are one of the most common types of accidents that cause injuries to non-motor drivers at signalized intersections. In order to analyze the effects of different non-motorized waiting areas on safety performance at signalized intersection, three typical non-motorized waiting areas (including common forms, advanced non-motorized parking line forms and BOX forms) were selected. The traffic conflict model based on the obtained data was used to analyze the relationship between traffic characteristics and conflicts caused by straight-through non-motor vehicles and right-turning motor vehicles. The chi-square test and the Kolmogorov-Smirnov test were used to determine the specific probability distribution of conflicting data, and a random parameter negative binomial (NB) model was developed. The results show that the random parameter NB model can overcome the defect that the traditional generalized linear model can’t describe the heterogeneity, and produce better fitting results than the fixed parameter NB model; The direct traffic of non-motor vehicles, right-turn traffic of motor vehicles, proportions of large cars and electric bicycles have significantly positive relationships with traffic conflicts., among them, the electric vehicle ratio has the largest elastic coefficient.

CICTP